Primary biliary cirrhosis (PBC) is a disease of unknown etiology characterized by slowly progressive intrahepatic cholestasis due to non-suppurative, presumably autoimmune, destruction of septal and the larger interlobular bile ducts. Because certain other autoimmune diseases appear to respond favorably to alkylating agents, a controlled trial of chlorambucil therapy for patients with PBC has been conducted. Twenty-four patients were admitted to this trial: 13 were randomized to receive chlorambucil therapy (0.5-4.0 mg/day) and 11 to the control (no treatment) group. The dose of chlorambucil was adjusted to reduce the peripheral blood lymphocyte count by 50% and maintain the polymorphonuclear leukocyte count above 1000 per c.mm. All patients have been followed for 2-5 years (mean = 43 months). During follow-up, two patients died: both were controls. Mean serum bilirubin levels remained almost constant in the treated group but increased by an average of about 50% each year in the controls. Mean serum albumin values increased slightly in treated patients but decreased in controls. Mean serum enzyme levels changed little in treated patients, but tended to rise in controls. Mean serum immunoglobulin (IgM and IgG) levels decreased to values within the normal range in all chlorambucil-treated patients, but did not change appreciably in controls. Liver biopsy histopathology after one and two years revealed less inflammation, slightly less fibrosis and less progression of the stage of disease in the treated than in the control patients. Potential side effects of chlorambucil therapy included the onset of menopause in two patients, localized herpes simplex or zoster in 3 and, in 4 patients, persistent leukopenia or thrombocytopenia requiring discontinuation of the drug. These findings strongly suggest that chlorambucil therapy retards the progression of primary biliary cirrhosis, and they provide an impetus to search for safer (e.g., noncarcinogenic) and more effective immunosuppressive regimes for the treatment of this disease. A large-scale double-blind, placebo-controlled trial of chlorambucil is necessary to assess the effect of this drug on survival of patients with PBC.